EP0218692A4

EP0218692A4 - Von willebrand factor.

Info

Publication number: EP0218692A4
Application number: EP19860902714
Authority: EP
Inventors: David Ginsburg; Stuart H Orkin; Randal J Kaufman
Original assignee: Childrens Medical Center Corp
Current assignee: Childrens Medical Center Corp
Priority date: 1985-04-11
Filing date: 1986-04-10
Publication date: 1988-03-22
Also published as: JPS62502589A; EP0218692A1; AU5776186A; WO1986006096A1; AU591671B2

Abstract

Von Willebrand's Factor (VWF) is produced using an expression vector that includes: 1) a DNA sequence encoding a functional VWF protein; and 2) regulatory DNA capable of effecting expression of that DNA sequence in a host cell transformed with the vector. Restriction fragment length polymorphisms (RFLP's) associated with the VWF gene are identified and used in a probe for determining the source of a VWF gene in a DNA sample. The gene for VWF is localized to the short arm of human chromosome 12 (12p).

Description

VON WILLEBRAND FACTOR

Background of the Invention This invention relates to mammalian Von Willebrand Factor (VWF), which is also known as Factor VIIIR, and to methods for obtaining and using VWF and for analyzing a mammalian DNA sample for the presence of VWF gene.

Blood coagulation in mammals involves the interaction of a number of protein factors and tissue components. One coagulation factor complex is called Factor VIII and is composed of at least two distinct proteins: Factor VIIIC (antihemophiliσ factor, the protein that corrects the coagulation disorder known as hemophilia A), and VWF. VWF is a protein that binds to platelets and is an essential component in platelet-vessel wall interaction during the clotting process. Zimmerman et al., Progress in Hematology Vol. XIII, "Factor VIII/Von Willebrand Factor" (Grune & Stratton 1983). Diminished or abnormal VWF activity can result in Von Willebrand's Disease (VWD) , a relatively common and complex hereditary bleeding disorder. The Factor VIII complex, obtained as a cryoprecipitate from donor blood, is administered as a therapy for VWD. Mitra, U.S. Patent 4,386,068. cDNA's specific for other coagulation proteins, e.g., Factor VIIIC, have been cloned and expressed in host systems [Wood et al (1984) Nature 312: 330-3371].

VWF is particulary difficult to analyze and produce because it is very large and there has been very little, if any, sequence data available heretofore. Moreover, cells that produce substantial amounts of VWF, e.g., endothelial cells and megakaryocytes, are difficult to grow in culture. Summary of the Invention

One aspect of the invention generally features producing VWF using an expression vector that includes: 1) a DNA sequence encoding a functional mammalian VWF protein; and 2) regulatory DNA capable of effecting expression of that DNA sequence in a host cell transformed with the vector. By a functional mammalian VWF protein we mean a protein that corresponds sufficiently to a naturally occurring mammalian VWF protein to have or to be processed to have the function of a naturally occurring Von Willebrand factor; processing may include glyσosylation and/or assembly into multimers. By a host cell we mean any suitable host cell such as a mammalian cell. At least some of the expression vector is exogenous to the VWF-encoding DNA sequence, meaning that it does not naturally occur in the same molecule with that sequence.

In preferred embodiments, the VWF is human VWF, and the host cell for producing the functional VWF is a eukaryotic cell, most preferably a mammalian cell.

A second aspect of the invention generally features analyzing a mammalian DNA sample using a probe comprising DNA encoding a VWF protein, or a fragment thereof, labeled with a detectable label. The probe is contacted with the sample to determine whether it hybridizes with the sample.

In a third aspect, the mammalian DNA that is to be analyzed is fragmented using restriction enzyme digestion, the DNA fragments are separated on the basis of size, the fragments are contacted with the above described probe, and the relative lengths of the DNA fragments that hybridize to the probe are determined. In preferred embodiments of both the second and third aspects, both the probe and the DNA being analyzed are human DNA.

The invention thus provides a relatively plentiful and pure source of VWF for treating bleeding disorders, for example, by administering: VWF to those with VWD; VWF in a stable complex with factor VIIIC to those with hemophilia A; or VWF to those with bleeding disorders associated with renal failure. The invention also provides diagnostic and research tools for evaluating VWF genes and defects in them using labled probes comprising the VWF gene or fragments of it. For example, a mammalian DNA sample can be analyzed using a DNA probe to detect a restriction fragment length polymorphism (RFLP) specifically associated with that VWF-related gene. By a VWF-related gene, we mean a normal VWF gene, or DNA characterized by the absence of part or all of the VWF gene or by a mutant VWF gene. Restriction fragment length polymorphism means an identifiable DNA sequence that is associated with the VWF-related gene and is conserved together with the VWF-related gene. RFLP's can be used to determine the pattern of inheritance of VWF genes, e.g. in a fetus at risk for one of the various forms of VWD. Other features and advantages of the invention will be apparent from the following descriptions of the preferred embodiments and from the claims.

Description of the Preferred Embodiment

We now describe preferred embodiments of the invention, first briefly describing the drawings.

I. Drawings Fig. 1A represents a restriction map of VWF DNA. Fig. IB represents a restriction map of the 3' portion of VWF cDNA derived from a plasmid designated pVWd and the 3' end of the overlapping clone pVWE6. Fig. 2 is a diagram of the chromosomal localization for the human VWF gene as determined by in situ hybridization.

II. Structure and Use

Section A below describes how to make functional human VWF by first assembling cDNA fragments to yield a cDNA coding for a functional human VWF protein and then expressing that cDNA in a host cell.

Section B describes the use of cDNA encoding VWF protein to analyze sample DNA, e.g., by detecting restriction fragment length polymorphisms (RFLP's) that are associated with the human gene. An assay can then be performed to determine the source of a VWF gene in a human DNA sample-e.g. which allele of each parent was inherited.

A. Functional VWF Protein The process of making functional VWF protein involves:

A) obtaining human mRNA; B) making a cDNA library from that mRNA and identifying cDNA fragments coding for VWF; C) assembling those fragments into a cDNA molecule encoding a functional VWF protein; and D) producing VWF by cloning that cDNA molecule into an expression vector that can be used to transform a eukaryotic host cell.

1. Obtaining the mRNA

As a source of VWF mRNA, a primary culture of human umbilical vein endothelial cells (HUVEC) is grown and passaged in cell culture in Medium 199 with 20% fetal bovine serum in the presence of bovine endothelial cell growth factor and fibronectin according to the method of Maciag et. al. [J. Cell Biol. 91:420-426 (1981), J. Cell Biol. 94.:511-520 (1982)]. Growth is markedly enhanced by the addition of heparin as described by Thornton et al. [Science 222: 623-625 (1983)]. To verify the presence of VWF mRNA, both the cultured cells and the conditioned medium are tested for the presence of VWF using anti-VWF antibody obtained by standard techniques. Standard immunofluorescense and ELISA assay, respectively, can be used for this purpose. After four additional passages, cells are harvested and total RNA prepared in guanidine HC1 by standard techniques.

2. Constructing a cDNA Library

Poly-A⁺ mRNA is isolated from total endothelial cell RNA by oligo-dT cellulose column chromatography. Two cDNA pools for the preparation of two different cDNA libraries are synthesized from the mRNA using standard techniques. For the first cDNA pool, oligo-dT is used as primer for the first strand synthesis, whereas for the second cDNA pool strand. The cDNA pools are made blunt-ended by treatment with T4 DNA polymerase and ligated to EcoRI linkers with T4 DNA ligase after protection of internal EcoRI sites by treatment with E. coli methylase. The linker-ligated cDNA's are then digested with an excess of EcoRI restriction enzyme and seperated from free linkers by passage over a Sepharose CL4B column.

The phage vector selected for carrying the VWF cDNA into a bacterial host is lambda gtll, a derivative of bacteriophage lambda which contains a bacterial gene for beta-galactosidase with a single EcoRI cloning site located near its 3' end, corresponding to the C-terminal portion of the beta-galactosidase protein. cDNA molecules are inserted into this site to construct a cDNA library.

By infecting an appropriate strain of bacteria with this phage using known techniques, a fusion protein will be produced containing most of beta-galactosidase at its amino terminus and a peptide fragment of the protein of interest at the carboxy-terminus. If this cDNA-encoded peptide contains one of the antigeniσ determinants of VWF, it is detected by screening with anti-VWF antibody. Large numbers of phage particles can be grown on a bacterial plate and the protein products transferred to a nitrocellulose filter and screened with a specific antibody to identify the location of the recombinant plaque producing the protein of interest.

Specifically, the VWF cDNA is ligated into EcoRI digested, phosphatase treated lambda gtll vector DNA and two libraries containing between 3-4 x 10⁶ recombinant clones each are plated and amplified. Nonrecombinant background as assessed by growth on ITPG/XGal plates is approximately 30%. Recombinant clones are obtained having cDNA inserts ranging in size from approximately 1 to 3 kilobases (kb) in length.

Affinity purified rabbit heteroantiserum prepared against human factor VIII-VWF is obtained using standard methods. The antiserum is passed over gelatin-sepharose, and adsorbed and eluted from a column of VWF-sepharose.

Recombinant clones from the above lambda-gtll endothelial cell cDNA libraries are screened as phage plaques in E. coli host strain Y1090 with this antibody at a 1:1000 dilution. Potential positive plaques are purified, replated and rescreened. For example, one primary filter screened with anti-human VWF antibody showed a positive plaque designated LVWd. As a positive control, purified VWF protein can be spotted onto the filter and can be detected at amounts between 100 and 0.1 nanograms (ng) total protein.

Positive plaques from the rescreening are purified and phage DNA prepared by standard methods. Purification and characterization of the cDNA insert in the above-described LVWd plague are described below.

The 553 bp cDNA insert of LVWd is purified by agarose gel eleσtrophoresis following EcoRI digestion and used as a probe to examine Northern blots of total mRNA from endothelial cells prepared as described above. Northern blot analysis was performed on total cell RNA from HPB-ALL (a T-cell line), endothelial cells (HUVEC) , fibroblasts, and Hela cells, with the LVWd cDNA insert as the hybridization probe. The LVWd cDNA probe hybridized with a single mRNA band between 8 and 10 kb in length. This mRNA species is large enough to code for a protein on the order of 250K dalton in molecular weight. This mRNA species was detected only in endothelial cells; no hybridization was observed with RNA's from the controls, i.e., human fibroblasts, Hela cells, or a human T-cell line (HPB-ALL). Thus the cDNA insert of clone LVWd corresponds to a segment of an mRNA molecule that is present only in endothelial cells and that is large enough to code for VWF. The clone contains a polypeptide epitope which reacts with affinity purified anti-VWF antibody.

The cDNA insert of LVWd is subcloned into plasmid pUC-13 (P-L Biochemicals) yielding the plasmid pVWd. Fig. 1A is a restriction map of cDNA coding for a VWF protein. Fig. 1B represents a restriction fragment map and sequencing strategy for DNA from the 3' portion of VWF cDNA derived from pVWd and the 3' end of the overlapping clone pVWE6 (see Fig. 1A). In both Fig. 1A and Fig. 1B, the DNA segment designated pVWD is the above-described cDNA fragment of the plasmid pVWD. Solid lines and arrows in Fig. 1B indicate regions sequenced by the method of Maxam and Gilbert [Meth. Enzymol.

65: 499-560 (1980)], with the solid circles indicating the end labeled restriction site. Dotted lines and arrows indicate regions sequenced by the method of Sanger et al. [Proσ. Natl. Acad. Sσi. U.S.A. 74: 5463-5467 (1977)] using PstI fragments subcloned into M13 mpll.

Table 1 shows the DNA sequence of the cDNA fragment of Fig. 1B. The cDNA sequence contains a single large open reading frame encoding 193 amino acids followed by a single in the same reading frame and orientation as the beta-galatosidase gene with which it was fused, consistent with expression of a fusion protein product. The predicted amino acid sequence for 193 amino acid residues at the carboxy-terminus of VWF is also shown using the standard single letter amino acid code, the six nucleotides at the beginning and at the end of the DNA sequence correspond to the synthetic EcoRi linker introduced by the cloning procedure. The single termination codon is marked by a diamond. Arrowheads indicate two potential N-glycosylation sites.

The chromosomal assignment of the VWF gene is established by the use of cDNA obtained as described above. Fig. 2 represents the hybridization patterns which indicate the localization of the VWF gene to the short arm of chromosome 12 (12p).

LVWd, pVWH5 and pVWH33 are on deposit with the American

Type Culture Collection (Rockville, Maryland, U.S.A.), under the respective accession numbers 53088, 53089 and 53090.

3. Assembling cDNA Fragments

In order to build a cDNA segment corresponding to a VWF protein, an insert from a positive plaque such as the 553 base pair insert of LVWd is used as a probe to rescreen the above-described HUVEC libraries, and to produce the restriction map of VWF cDNA shown in Fig. 1A.

Specifically, the cDNA insert is purified in low melt agarose (Bethesda Research Labs) following EcoRI digestion.

and subcloned into the EcoRI digested and phosphatased pUC-13 plasmid (P-L Biochemicals) to yield pVWD (see Fig. 1A). The EcoRI insert is radiolabeled and used as a probe to rescreen the HUVEC cDNA library. The positive recombinant phage are purified and subcloned into pUC-13. Two such recombinants, pVWE2 and pVWE6, are illustrated in Fig. 1A. The restriction map is deduced by standard methods. A 217 bp EcoRI/PstI fragment from the 5' end of pVWE2 is used to rescreen the HUVEC cDNA library and a third series of overlapping VWF cDNA clones is identified, one of which, pVWG8b, is shown in Fig. 1A. Using the same methods, a fourth series of overlapping clones is obtained, including pVWH33 and pVWH5 which, together with the sequence shown in Table 1, span DNA sufficient to encode an entire monomer of human VWF protein. By using the appropriate restriction enzymes, reagents, and procedures, one skilled in the art can ligate together cDNA fragments of pVWH33, pVWH5, and the sequence of Table 1, for example, to construct a cDNA encoding an entire monomer of VWF protein.

Construction and Expression of Full-Length cDNA

More specifically, cDNA clones pVWH33, pVWH5 and pVWE6, which span 9kb pairs of DNA and encompass the entire protein coding region of VWF, were selected to construct full length cDNA. Nucleotide sequence of the ends of these clones confirmed that together they include the translational start, protein coding and translational stop sequences. The full length cDNA was constructed by standard techniques using fragments derived from individual clones. A fragment of pVWH33 from the left hand EcoRI site to the unique BamHI site was linked to a pVWH5 fragment from the BamHI site to the Sac I site. This was then linked to a fragment from pVWE6 from the Sac I site to the right hand EcoRI site. The full-length cDNA was then inserted as an EcoRI fragment into the EcoRI site of the expression vector pMT2 in which transcription occurs under the control of the adenovirus major late promoter.

pMT2 is a derivative of the mammalian cell expression vector p91023(B) (Wong et al., 1985, Science 228;810) in which the tetracycline resistance marker is substituted for the ampicillin resistance marker. The functional elements of the VWF expression plasmid have been previously described (Kaufman, 1985, Proc. Natl. Acad. Sci. USA82:689). pMT2-VWF contains the SV40 origin and enhancer element; the adenovirus major late promoter with the first, second and two thirds of the third tripartite leader; an intron from the 5'splice site from the first late leader of adenovirus and 3' splice site from an immunoglobulin gene (Kaufman & Sharp '82); the VWF cDNA; DHFR coding region, and SV40 early polyadenylation site; the adenovirus VA genes in a derivative of pBR322 containing the Col El origin of replication and ampicillin resistance.

pMT2-VWF was grown in E. coli DH5 in order to prevent deletion of the VWF sequence. Plasmid DNA was prepared by twice banding to equilibrium on CsCl gradients.

4. Production and Use of VWF

The above-described cDNA encoding a VWF protein can be inserted into a suitable vector and expressed in any one of a number of mammalian expression systems known to the art, for example using the general method described by Wood et al. (1984) Nature 312;330-337. The resulting product with any necessary post-translational processing, yields a mature Von Willebrand factor. Host systems can be selected for appropriate post translational processing of the VWF gene product, and enable efficient recovery of VWF. Active VWF has thus been expressed in COS monkey cells and in CHO cells, for example, as described below. Pure VWF produced in this way will be useful in the treatment of VWD, and patients with chronic renal failure whose abnormal bleeding times are corrected by crude cryoprecipitate. Pure VWF can also be used to carry, stabilize, and improve the therapeutic efficacy of factor VIII:C.

Expression of VWF in Monkey COS Cells.

The SV40-transformed COS monkey cells (clone M6) have been described (Horowitz et al . , 1983 , J. Mol. Appl. Genet. 2 : 147-149). DNA transfections using pMT2 and pMT2-VWF were performed as described (Kaufman, 1985, Proσ. Natl. Acad. Sci. USA 82:689 ) by the. DEAE-dextran procedure with the addition of a chloroquin treatment (Sompayrac and Danna, 1981, Proc. Natl. Acad. Sci. 78:75757578; Luthman and Magnusson, 1983, Nuc. Acids Res. 11:12951308). Transfected cells were fed with DMEM (Dulbecco's Modified Eagle's Media) with 10% total bovine serum for 48 hr. Then the media was removed, the cells rinsed, and serum-free DMEM applied (4 ml per 3x10⁶ cells) for measurement for VWF using an inhibition ELISA assay in which purified VWF was adsorbed onto the surface of microtiter wells followed by anti-VWF antibody. The ability of test material to displace anti-VWF antibody from the immobilized antigen was tested using peroxidase conjugated anti-rabbit IgG as the indicator substance. Media from COS cells transfected with expression vector pMT2-VWF containing VWF cDNA produced between 50 and 300 ng/ml VWF antigen in three seperate transfections. COS cells transfected with vector pMT2 alone did not produce any protein reacting in the ELISA assay.

Processing of Recombinant VWF

A transfection of COS cells was performed as above and 72 hours post-transfection, the media was replaced with fresh cysteine-free media containing ³⁵S-cysteine. After an additional 1 to 5 hours of incubation, the media was removed and cell extracts were prepared as described (Kaufman and Sharp, 1982, J. Mol. Biol. 159:601). The cell extracts and media were then used for studies of VWF processing and multimer assembly. VWF was immunoprecipated by incubation with rabbit anti-human VWF antibody, followed by protein-A sepharose. The immunoprecipated material was washed in a buffer containing 0.1% SDS and Nonidet P40 to minimize non-specific adsorption of other proteins to the immune complex. The precipitated proteins were then analyzed by 4 to 6% SDS-PAGE in the presence and absence of lmM DTT. Recombinant cell lysates contained a band migrating with an apparent molecular weight of 260 kd. Cell media contained a mixture of 260 kd and 220 kd species. Immunopreσiptates derived from COS cells or media transfected with non-recombinant plasmid did not contain these two bands. Analysis of non-reduced species by 4% SDS-PAGE showed a series of four or five very high molecular weight proteins varying from 1 to 3 or 4 million daltons.

Biological Activity of Recombinant VWF

COS cells were transfected as above and 48 hr. post- transfection were rinsed and fed with serum free media. Two hundred ml of serum free COS cell media was collected after incubation with COS cells for 24 hours. It contained between 50 and 200 ng/ml VWF by ELISA. The serum free media was concentrated by dialysis against 50% Ficoll to a concentration of 2.2 ug/ml VWF protein for use in the competitive binding assays.

This recombinant VWF was used in varying concentrations as a competing ligand against purified, radiolabelled human VWF multimers in assays as described by Luscalzo, J. and

Handin, R.I., 1984, Biochem. 23:3880. A concentration of lug/ml competed for 50% of the collagen binding (I.C. 50). This was ten fold less than the I.C. 50 when purified human VWF was used as competing ligand. COS media from cells transfected with pMT2 plasmid alone did not compete for collagen binding sites. Similarly, the I.C. 50 for recombinant VWF binding to platelet glycoprotein lb was 2 ug/ml when using freshly isolated platelets, and 5 ug/ml when using formalin fixed platelets as the source of receptor. These values are identical to those obtained with purified human VWF. Again, media form COS cells transfected with nonrecombinant plasmid did not compete for binding. These results demonstrate that COS cell derived VWF is functional as determined by collagen and platelet binding.

VWF Expression in CHO Cells

Although a number of systems are available for the production of VWF in mammalian cells, one particularly useful approach to obtain high level expression is to select for cells that contain a high degree of amplification of the heterologous VWF gene. One amplifiable marker which is available for this approach is the dihydrofolate reductase gene for which cells harbouring increased gene copies can be selected by propagation in increasing concentrations of methotrexate (MTX) (Kaufman and Sharp, 1982, J. Mol. Biol. 159: 601). This approach can be used to select and amplify the VWF gene in a variety of different cell types and has been used to obtain expression of active, full-length human VWF in Chinese hamster ovary cells.

Coamplification and Coexpression of VWF and DHFR in DHFR Deficient Chinese Hamster Ovary (CHO) Cells

The VWF expression plasmid pMT2-VWF and the DHFR expression plasmid pAdD26SV (A) 3 (Kaufman and Sharp, 1982, Mol. Cell.Biol 2: 1304) were introduced into DHFR deficient CHO DUKX-BII cells by calcium-phosphate coprecipitation and transfection. DHFR+ transformants were selected for growth in alpha media with dialyzed fetal calf serum and subsequently selected for amplification by growth in increasing concentrations of MTX (sequential steps in 0.02, 0.2, 1.0, and 5.0 uM) as described (Kaufman et al., 1985, Mol. Cell. Biol. 5:1750). One transformant, designated XMTVWF-C1, was isolated in alpha media and propagated in MTX as described above. The expression of VWF, monitored by ELISA, as a function of increasing levels of MTX resistance is shown below. uM [MTX] ng/ml/day

0.2 56

1.0 91

5.0 278

VWF expression increased with increasing levels of MTX resistance.

The VWF derived from the CHO cells was assayed by a direct ELISA assay using a rabbit anti-human VWF antibody (Calbiochem) immobilized on microtiter plates and a secondary antibody conjugated to horse raddish peroxidase (Dacco). There was minimal activity in the media from the original CHO cells (less than 5 ng/ml). Values were determined by comparison to a standard derived from normal human pooled plasma (1 unit/ml) which was assumed to contain 10ug/ml of VWF. VWF expression has also been verified by ³⁵S-cysteine labeling of the cells and analysis of the conditioned media and cell extracts by immunopreσipitation with rabbit anti-human-VWF antisera (Calbiochem) and electrophoresis of SDS-polyacrylamide gels as described above for the VWF derived from transfected COS cells.

5. Nucleotide sequence of VWF rDNA

The VWF DNA sequence was derived from the same overlapping cDNA clones which were used in the construction of the full-length expressed clone pMT2-VWF. An additonal 70 bp of 5' untranslated region, derived from the most 5' clone isolated, pVWK7, has been included. The entire sequence was determined on both strands using the Sanger dideoxy method on single-stranded M13 subclones (Sanger, F. Nicklen., S. and Coulson, A.R. , 1977, Proc. Natl. Acad.- Sci. U.S.A. 74: 5463-5467; Messing, J., 1983, J. Methods Enzymol. 101:30-78) . Sub-clones for sequencing were generated by exonuclease digestion of the inserts of cDNA clones pVWH33, pVWH5, and pVWE2, using nuclease Bal31 or T4 DNA polymerase. Gaps were completed by subcloning appropriate restriction fragments from the same clones into M13 mp10 and mpll.

The sequence of 8588 base pairs is shown in Table 2. It contains continuous open reading frame encoding a polypeptide of 2815 amino acids.

There are three lines of evidence supporting the authenticity of the indicated translational start site. First, there is an upstream nonsense codon in the major open reading frame. Second, the only other upstream start codon is followed almost immediately by an in frame stop codon. Finally, the presumptive initiator methionine is followed by a classical signal peptide sequence, as expected from this secreted glycoprotein.

In order to characterize the 5' untranslated region, several other clones containing the 5' segments of the VWF cDNA partially sequenced but no two independent clones were found to have the same 5' end. As shown here pVWK7 extended the farthest 5' .

The apparent descrepancy between the length needed to encode an estimated 260 kd VWF presursor (about 7kb) and the observed VWF message size of 8-9 kb has been previously noted, and has led some investigators to postulate the presence of an extremely long 5' untranslated region (Lynch et al., 1985, Cell 41:49). The presence of an 8.3 kb continuous open reading frame clearly shows that the primary VWF transcript is much larger than the 260 kd suggested by SDS-PAGE. The predicted molecular weight is approximately

300 kd. , even before the extra contribution from glycosylation is taken into account. SDS-polacrylamide gels are known to be an inacccurate way to estimate molecular weights in this range, althouugh it has not been formally excluded that multi-step processing occurs, with rapid formation of a relatively stable 260 kd intermediate. Since the VWF "pro-piece" can be found intact in the circulation as a 100 kd glycoprotein, multi-stage processing seems unlikely and the pro-VWF is clearly larger than previously suspected and processed in a single step.

The function, if any, of the large VWF pro-piece, is currently unknown. Its fate after trans-membrane secretion is also not fully clear. However, the identity of the VWF pro-piece with a 10Kd plasma glycoprotein has recently been established so that at least some of the intact pro-piece leaves the cell. The N-terminal sequence of the 100Kd glycoprotein which corresponds to the predicted sequence of the propolypeptide implies that signal peptide cleavage occurs at the position shown by the arrow in Table 2. This agrees with the consensus sequences usually involved in signal peptidase cleavage (Watson, Nuc. Acids Res., 12, 1984).

Sadler et al., 1985, Proc. Natl. Acad. Sci. USA 82:6394 have recently published a partial VWF sequence, and noted the presence of repeated elements. Analysis of the complete nucleotide sequence, however, reveals much more extensive repetition in the VWF structure than shown by previous data. We confirm the presence of three complete copies of a repeat labelled "domain A" by Sadler et al. We have retained the terminology for this repeat and have confirmed the homology of the 5' end of its first copy, missing in the clones of Sadler et al. A striking feature, not noted by those authors, is the paucity of cys residues within this region, which occupies about 600 amino-acids in the center of the VWF sequence. In contrast, the regions at each end of the molecule, from nucleotides 208 to 3833, and 5729 to 8582 in our sequence, are extremely rich in Cys, and account almost entirely for the high Cys content of VWF. Cys is in facat the most abundant amino-acid in the prepro-VWF sequence, accounting for 8.3% of the residues. In the region outside "domain A", Cys accounts for 10.4% of the amino acids.

Further analysis of these Cys rich regions indicates that they too can be arranged as a series of 6 repeats of a basic 400 amino acid unit. The first repeat block begins almost immediatly after the signal peptide. The second repeat is truncated just preceding this duplication contains the cleavage site between the pro-piece and mature VWF. Following the third repeat, the sequence is interrupted by the triplicated cys-poor "A domain" repeat, after which the Cys rich repeats resume encoding the C-terminal end of the molecule. The fifth and sixth repeats are incomplete, but include the region of the short "B domain" repeats of Sadler et al. Those authors' "C domain" repeats follow in two copies. Since no free sulfhydryl groups can be detected in multimeriσ VWF, all these Cys residues are involved in interchain and intrachain disulfide bonds which are important determinants of the tertiary and quaternary structure of the protein.

B. Analyzing Sample DNA

DNA sequence polymorphisms are neutral variations in DNA sequence present throughout the genome, which can often be detected by restriction enzyme digestion and blot hybridization analysis. By neutral we mean that the variations per se are not themselves responsible for any phenotypic trait. However, the value of polymorphisms is that they are linked to or associated with adjacent portions of the genome, and therefore they can be used as markers of those portions of the genome.

Two types of DNA sequence polymorphisms have been described. One type involves single nucleotide changes, or small insertions or deletions, which result in the presence or absence of a particular restriction enzyme recognition site. In another type of polymorphism, a large segment of DNA of unknown function varies widely in length among individuals. Both types of sequence differences are inherited in Mendelian fashion.

RFLP's linked to the VWF gene are identified in genomic DNA from individuals examined by cutting sample DNA with a series of restriction enzymes. The resulting restriction length fragments are segregated by molecular weight. Hybridization with a radiolabeled VWF cDNA probe, e.g., the cDNA insert from clone pVWE6 yields (e.g., using Southern blot techniques) a unique band pattern. For example, RFLP's are detected by the above procedure using restriction enzymes TaqI and Saσl and the pVWE6 probe. Specifically, peripheral blood specimens are collected in 10% acid-citrate-dextrose. High molecular weight DNA is prepared by standard techniques either from dextran sedimented leukocytes or isolated nuclei separated by centrifugation following Triton X-100 solubulization. From 2 to 16 g of DNA are digested to completion with the restriction enzymes TaqI and Sad. DNA fragments are then fraσtioned by electrophoresis in 0.6 to 1.0% agarose gels and transferred to nitrocellulose filters. The probes constructed from pVWE6 are labeled and hybridized with the DNA on the above-described nitrocelloulose filters. The hybridized filters are washed and examined by autoradiography. Once identified RFLP's can be used as described above to assay a sample of human DNA and determine the source of a VWF gene in that sample. For example, DNA of parents and other family members can be screened, e.g., by Southern blotting, with such a probe, and then fetal DNA can be screened to determine the inheritance pattern of its VWF alleles.

Other Embodiments

Other embodiments are within the following claims. For example, the techniques described apply to other mammalian systems. Other RFLP's can be used. Hybridization probes can be used for other assays or research techniques.

Claims

AMENDED CLAIMS

[received by the International Bureau on 8 September 1986 (08.09.86); original claims 1-16 replaced by new claims 1-21 (2 pages)]

1. A cDNA sequence encoding functional VWF which substantially comprises the VWF cDNA sequence of at least one of ATCC Nos. 53088, 53089 and 53090.

2. A DNA sequence of claim 1 which is further characterized by encoding the protein sequence substantially as depicted in Table 2.

3. A vector for transforming a host cell, said vector comprising a DNA sequence of claim 1 and regulatory DNA capable of effecting expression of said DNA sequence in the host cell to produce functional VWF, at least a portion of said vector being exogeneous to said DNA sequence.

4. A host cell transformed with the vector of claim 3.

5. A transformed host cell of claim 4 wherein the host cell is a eukaryotic host cell.

6. A transformed host cell of claim 4, wherein the host cell is a mammalian host cell.

7. A method for producing functional human VWF which comprises culturing a transformed host cell of claim 4.

8. A method for producing functional human VWF which comprises culturing a transformed host cell of claim 5.

9. A method for producing functional human VWF which comprises culturing a transformed host cell of claim 6.

10. The method of claim 7 which further comprises purifying the VWF so produced.

11. The method of claim 8 which further comprises purifying the VWF so produced.

12. The method of claim 9 which further comprises purifying the VWF so produced.

13. VWF produced by the method of claim 7.

14. VWF produced by the method of claim 8.

15. VWF produced by the method of claim 9.

16. A DNA probe for analyzing a mammalian DNA sample, said probe comprising DNA encoding VWF or a fragment thereof, labeled with a detectable label.

17. The probe of claim 16, wherein said DNA encoding VWF os human DNA.

18. A method of analyzing a mammalian DNA sample comprising contacting said DNA with the probe of claim 17 and determining whether said probe hybridizes with said sample.

19. The method of claim 18, wherein said DNA sample is human DNA.

20. A method of analyzing a mammalian DNA sample comprising: fragmenting said DNA sample by restriction enzyme digestion, separating said DNA fragments on the basis of size, contacting said fragments with the probe of claim 17 and determining the relative lengths of said DNA fragments.

21. The method of claim 20, wherein said DNA sample is human DNA. STATEMENT UNDER ARTICLE 19

In response to the 7 July 1986 Search Report, Applicant submits herewith replacement sheets for pages 41 and 42 of the application as filed. New claims, believed to contain no new matter, have been substituted on the replacement pages for original claims 1-14. Of the new claims, claims 16-21 are identical except for numbering to originally presented claims 11-16.